Raspberry Pi 2 Launches With Quad-Core ARM SoC

[duby229]

A bit off topic perhaps, but I can't help myself - I actually wish developers tried to stay the hell away from stellar hardware. If the madmen that developed visualstudio actually used 900Mhz machines and sticked to them you would probably get something useful that worked and was efficient instead of frameworks running on top of other frameworks that require a gazillion ultrabytes ram and a yottahertz processor just to open a file dialog with a new fancy design that is pointless and waste screen space compared to an old open file dialog that was compact and did essentially 100% the same job only more efficient and intuitive! AAAAAAAARGH!!!
...sorry but I just HAD to get that out of the system...

Absolutely agree,

Thin and lite is -always- better 100% of the time.

To be fair: I think MS actually did start to learn that lesson somewhat. They might still be pigheaded about it, but they know.

[MartinN]

So, finally on par with cheap chinese boards in terms price to performance ratio.

I like how you say "cheap Chinese" as if the hardware you're using to write your reply on wasn't manufactured 100% in China .

[Veto]

Having both the ODROID-C1 and the RPi, I can address the Ethernet.
...
Given a choice, I would not have bought the C1 today.

Thank you! This message is probably the most informative and made it worthwhile wading through this troll-infested thread.

I actually strongly considered buying the C1, in the hopes of better performance and maybe Ubuntu support. I had some reservations though with regard to actual support - and your post confirms this. The RaspberryPi may not be the most performant board out there - but its support, community and documentation trumps the competitors. I will order a RPi 2 instead, now it is available :-)

[curaga]

Ordered one. Finally the price-to-perf ratio is acceptable, and it coincides with Eric's open vc4 work. This may just become the nicest ARM board for a year or two GPU-wise (likely usb and networking still suck, so no servers/NAS use).

So why not a chinese one? A similar price in name may end up with 30$ more due to shipping, plus way less support. I ended up getting mine for 33$, even less than the official price, including shipping (group order).

[Alliancemd]

Re

And just yesterday I made a command for some Odroid boards from China...
I can buy a RPi from where I live but I thought, to get them from China will be a bit more expensive but at least those are a lot more powerful...
Now these news...

[blackshard]

Programming 700MHz CPU in python? Sounds like if those who do it will learn to be masochistically inclined. It is slow on its own and programming it on python would make it painfully slow. Sure, sometimes you do not need much performance. But when you do, it going to be real nightmare.

Python is a good choice when you don't need performance. But may be also a good choice when you NEED performance, because it can be interfaced with dynamic libraries written in compiled languages. I am absolutely not a Python adept, but importing external libraries (even system libraries as is) in python is sensationally easy.

[Baconmon]

I use kodi on RPi1 currently..

On RPi2, will I be able to watch h265 movies now?.. And how about daala videos, or VP9?..

After you tell me "No", then I'd also like to ask: Would it be possible for it to run some of those codecs some day if GPU decoding was added to it for some of those codecs?.. Or will I just have to keep looking for h264 versions of shows/movies for eternity?..

[curaga]

For the Nth time, GPUs are not suited for generic video decoding accel. Most codecs do not parallelize.

[SystemCrasher]

I like how you say "cheap Chinese" as if the hardware you're using to write your reply on wasn't manufactured 100% in China .

Actually, I can also type reply from chinese board, if desired. I can also use keyboard, mouse and LCD made in china to make it more funny. All major components of board I have were made in china and I can admit they have got reasonable qualty and parameters. Sure, they licensed core from ARM. But they also developed own cores, based on MIPS designs and as far as I understand, they technically have full IP on these cores and can redesign them at will. So, technically china seems to be able to be fully technologically independent, should they need or want this. So I doubt you can troll them about stuff like this. I think China can be 100% independent in terms of digital technologies, should they need it for whatever reasons. And actually it is china who manufactures most of stuff around.

Python is a good choice when you don't need performance. But may be also a good choice when you NEED performance, because it can be interfaced with dynamic libraries written in compiled languages. I am absolutely not a Python adept, but importing external libraries (even system libraries as is) in python is sensationally easy.

As I can see, python usually used by those who would use BASIC some years ago. It also haves weird syntax which isn't anyhow resembles languages used in libs and building own libs would require to learn something completely different. So once python nuts need some speed in THEIR algos, they're mostly doomed. Speking for myself I dislike python for awkward syntax which was definitely targeted at dullards (who else would to learn to format programs properly through mental rape at syntax level?) and for breaking compatibility each and every time. So every couple of years or so most of python code turns into useless garbage unless someone willing to rewrite half of program. So good luck to run python 2.4 code these days. Yet, somehow, I can build 1990 year program with GCC and it will work. That's what I call long term compatibility. Sure, maybe someone happy to rewrite code every couple years just because some super-duper language devs decided so. But that's not about me for sure. I really do not get why there is so much buzz about this crap. And programming 700MHz CPU using slow and lame interpreter sounds like painful thing in terms of performance and overall resources consumption.

[duby229]

For the Nth time, GPUs are not suited for generic video decoding accel. Most codecs do not parallelize.

My understanding is that's only because they are written in languages like C. And C programmers can't get over what they spent too many years learning. What codecs do is inherently parallelizable.